Animal studies and epidemiological studies have suggested that Ca supplementation (with Ca supplements or dairy products) may be associated with weight loss in human adults. We aimed toassess whether any association was present by reviewing relevant randomized controlled trials in human subjects. The study was a systematic review and subsequent meta-analysis of randomized controlled trials that used Ca supplementation as an intervention in persons 18 or more years of age, and that reported body weight as a final outcome. A total of thirteen randomized controlled trials were included in the meta-analysis. There was no association between the increased consumption of either Ca supplements or dairy products and weight loss after adjusting fordifferences in baseline weights between the control and intervention groups (P=0·19 and 0·85, respectively). We therefore concluded that Ca supplementation has no statistically significant association with a reduction in body weight.

Genistein and parathyroid hormone (PTH) are anabolic agents that stimulate bone formation through their direct actions in osteoblastic cells. In the present study, we aimed to determinewhether genistein modulates the actions of PTH in human osteoblastic SaOS-2 cells in an oestrogen-depleted condition. The present results showed that genistein (10−8 to 10−6 m) induced alkaline phosphatase (ALP) activity and osteoprotegrin (OPG) expression in SaOS-2 cells in a dose-dependent manner. These effects could be completely abolished by co-treatment with oestrogen antagonist ICI 182780 (7α-[9-[(4,4,5,5,5-pentafluoropentyl)sulfonyl]nonyl]-estra-1,3,5(10)-triene-3,17β-diol). Genistein (at 1μm) could stimulate the mRNA expression of receptor activator of NF-κB ligand (RANKL). As OPG and RANKL are known to modulate osteoclastogenesis, the ability of genistein to modulate OPG and RANKL expression in SaOS-2 cells suggested that it might modulate osteoclastogenesis through its direct actions on osteoblastic cells. PTH (at 10nm) stimulated ALP activity, induced RANKL mRNA expression and suppressed OPG mRNA expression in SaOS-2 cells, confirming its bi-directional effects on osteoblastic cells. Pre-treatment of SaOS-2 cells with genistein andoestrogen not only enhanced PTH-induced ALP activity, but also attenuated PTH up regulation ofRANKL mRNA expression and PTH down regulation of OPG mRNA expression. Taken together, the present study provides the first evidence that genistein could modulate the actions of PTH in human osteoblastic SaOS-2 cells in an oestrogen-depleted condition.

We have recently reported that intake of soya protein isolate (SPI) inhibited the DNA-binding activities of hepatic thyroid hormone receptor (TR). The genes for acetyl-CoA carboxylase (ACC), a rate-limiting enzyme in fatty acid synthesis, contain the thyroid hormone response element in their promoters and are regulated by TR. The present study has examined the effect oflong-term feeding of SPI and soya isoflavones (ISF) on the gene expression and protein phosphorylation of different ACC isoforms in different tissues and plasma triacylglycerol (TAG) levels in rats. Sprague-Dawley female rats were fed diets containing 20% casein or alcohol-washed SPI with or without supplemental ISF for 70, 190 and 310d. SPI intake significantly reduced plasma TAG concentrations compared with casein, whereas supplemental ISF had no effect. Hepatic ACCα and ACCβ mRNA abundance and protein content were markedly lowerin the rats fed SPI than in those fed casein. The protein contents of ACCα in the kidneyand ACCβ, the predominant isoform in the heart and kidney, were unchanged by dietary SPI.The ratios of phospho-ACCα/ACCα and phospho-ACCβ/ACCβ were not differentamong dietary groups in all tissues measured. The present study demonstrates that ingestion ofSPI decreases plasma TAG level and down-regulates ACCα and ACCβ gene expression in the liver but not in the heart and kidney. The results indicate that the effect of SPI is tissue-specific and that alteration of ACC gene expression rather than phosphorylation status may play a major role in the regulation of ACC activities by soya proteins.

The diets of laboratory rats were isotopically and nutritionally manipulated using purifiedC3 and/or C4 macronutrients to investigate the routing of dietary carbonto bone collagen biosynthesis. Diets were formulated with purified proteins, carbohydrates andlipids of defined composition and natural abundance stable isotope ratios. Bulk protein and constituent amino acid δ13C values determined for whole diet and bone collagen provided the basis for assessing isotopic fractionation and estimating the degree of routing versus synthesis de novo of essential, non-essential and conditionally indispensable amino acids. Essential and conditionally indispensable amino acids were shown to be routed from diet to collagen with little isotopic fractionation whereas non-essential amino acids differed by up to 20‰. Mathematical modelling of the relationships between macronutrient and tissue δ13C values provided qualitative and quantitative insights into the metabolic and energetic controls on bone collagen biosynthesis. Essential amino acids comprise 21·7% of the carbon in collagen, defining the minimum amount of dietary carbon routing. Estimates of 42 and 28% routing were shown for the non-essential amino acids, glycine and aspartate, respectively. In total, the routing of non-essential and conditionally indispensable amino acids was estimated to equal 29·6% of the carbon in collagen. When the contribution of carbon from the essential amino acids is also considered, we arrive at an overall minimum estimate of 51·3% routing of dietary amino acid carbon into bone collagen.

Annealing and heat-moisture treatment (HMT) are shown to be suitable methods to increase the yield of resistant starch type 3 (RS3) from Novelose 330 by up to 75%. Peak temperatures of approximately 121°C were used to produce to a sufficiently high thermal stability of the hydrothermal modified RS3 products for a wide range of applications. HMT significantly increased the crystallinity up to 40%. An in vivo feeding experiment with Wistar rats showed that fermentation of Novelose 330 dominated in the proximal colon, but degradation of HMT-Novelose was more dominant in the distal colon, leading to higher butyrate concentrations in this segment of the large bowel. Large-bowel surface and crypt length increased in the proximal colon in rats fed the Novelose 330-containing diet. In contrast, after the intake of HMT-Novelose, maximal values were found in the distal segment. The lower pH and higher butyrate concentration of the caecal and colonic contents significantlysuppressed the formation of secondary bile acids in RS3-fed rats. The formation of secondary bile acids was inhibited more strongly by HMT-Novelose than by Novelose 330. The Ki-67-immunopositive epithelial cells in the colon of RS3-fed rats indicated the establishment of an optimalbalance in the dynamic process of mucosal regeneration. HMT provides a method for the economical production of a high-quality RS3 with dominating prebiotic properties in the distal colon for health-promoting applications.

The intestine of newborn pigs develops rapidly during the first days postpartum. We investigated if feeding milk replacer (infant formula) as an alternative to colostrum has compromising effects on nutrient digestive function in the neonatal period. Nineteen piglets born at termwere assigned to one of four treatments: (1) newborn controls; (2) natural suckling for 24h; (3) tube-fed formula for 24h; (4) tube-fed porcine colostrum for 24h. All three fed groups showed significant increases in small-intestinal and colonic weights, villous heights and widths, maltase and aminopeptidase A activities, and decreases in dipeptidylpeptidase IV activity, relative to newborn pigs. Following oral boluses of mannitol, lactose or galactose, formula-fed pigs showed significantly reduced plasma levels of mannitol and galactose compared with colostrum-fed pigs. Activity of intestinal inducible NO synthase and plasma levels of cortisol were significantly increased, whereas intestinal constitutive NO synthase and α-tocopherol were decreased in formula-fed pigs compared with colostrum-fed pigs. Although formula-fed pigs only showed minor clinical signs of intestinal dysfunction and showed similar intestinal trophic responses just after birth, as those fed colostrum, lactose digestivecapacity was markedly reduced. We conclude that formula-feeding may exert detrimental effects on intestinal function in neonates. Formula-induced subclinical malfunction of the gut in pigsborn at term was associated with altered NO synthase activity and antioxidative capacity.

Factorial approaches to estimate energy requirements of growing pigs require estimation of maintenance energy requirements. Heat production (HP) during fasting (FHP) may provide an estimate of maintenance energy requirements. Six barrows were used to determine effects of feedinglevel on components of HP, including extrapolated plateau HP following a 24h fast (FHPp). Based on a cross-over design, each pig was exposed to three feeding levels (1·55, 2·05 and 2·54MJ metabolisable energy/kg body weight (BW)0·60 per d) between 30 and 90kg BW. Following a 14d adaptation period, HP wasestimated using indirect calorimetry on pigs housed individually. Dynamics of HP were recordedin pigs for 5d during the fed state and during a subsequent 24h fast. Metabolisable energy intake was partitioned between thermal effect of feeding (HPf), activity HP (HPa), FHPp and energy retention. Feeding level influenced (P<0·05) total HP during the fed state, HPf and activity-free FHPp (609, 644 and 729 (se 31) kJ/kg BW0·60 per d for low, medium and high ME intakes, respectively). The value of FHPp when expressed per kg BW0·60 did not differ (P=0·34) between the three subsequent experimental periods. Feeding level did not (P=0·75) influence HPa. Regression of total HP during the fed state to zero metabolisable energy intake yielded a value of 489 (se 69) kJ/kg BW0·60 per d, which is a lower estimate ofmaintenanceenergy requirement than FHPp. Duration of adaptation of pigs to changes in feeding level and calculation methods should be considered when measuring or estimating FHPp, maintenance energy requirements and diet net energy content.

Elevated plasma homocysteine is a risk factor for atherosclerotic disease. In the present study, we have examined whether the oxidative stress due to a low level of vitamin B6 accelerates the development of homocysteine-induced atherosclerosis in rats. First, the effect of homocysteine thiolactone intake (50mg/kgperd) on vascular integrity, lipid peroxide concentration, endothelial NO synthase (eNOS) expression and biochemical profiles was examined at day 1, day 21 and day 42 (five rats per group). The histochemical staining ofthe rat aorta showed no change at day 1 and day 21, but the subendothelial space was observed to be enlarged in rat aorta at day 42 with exposure to homocysteine thiolactone. Expression of eNOS was observed in rat aorta at day 42, but not at day 1 and day 21. Serum lipid peroxide concentration and biochemical profiles including glucose cholesterol and triacylglycerol showed no change at any day. Second, the effect of homocysteine thiolactone intake in the presence and absence of vitamin B6 on vascular integrity was examined at day 1 and day 14 (five rats per group). Aortic lesions were observed in vitamin B6-deficient rat aorta at day 14 but not in vitamin B6-supplemented rats. The expression of eNOS was also observedin vitamin B6-deficient rat aorta at day 14. Serum lipid concentrations of the vitamin B6-deficient group significantly increased compared with concentrations of the vitamin B6-supplemented group, though serum concentration of homocysteine did not change between both groups. These results suggest that the oxidative stress caused by a low level of vitamin B6 accelerates the development of homocysteine-induced atherosclerosis in rats.

The present study consisted of two experiments. The first experiment was conducted to determine the effects of lipoic acid (ALA; 200mg/kg) on plasma lipids and insulin sensitivity of whole-body tissue in broilers treated with or without corticosterone (5mg/kg). Chickens received these agents from 2 to 5 weeks of age in a 2×2 factorial arrangement. Thereafter, from 39 to 42d of age, insulin sensitivity was estimated using the euglycaemic and hyperinsulinaemic clamp technique. Experiment 2 examined whether ALA supplementation for 5 weeks (400mg/kg) would alter short-chain acyl-CoA concentration in the liver and in vitro lipolysis of an adipose tissue slice, in relation to noradrenaline (10μm) supplementation. In experiment 1, ALA had no effect on the corticosterone-induced negative growth performance. ALA lowered plasma glucose level (P<0·05) and, in contrast, increased triacylglycerol level (P<0·05). These responses to ALA had, however, no interrelationship with corticosterone. The rate of glucose uptake of whole-body P<0·05) but was not affected by noradrenaline supplementation. This study suggests that ALA stimulates the insulin sensitivity of tissues regardless of corticosterone-dependent metabolism and that the ALA-induced fatty acid metabolism of broilers differs between the liver and adipose tissue.

The nutritional composition of the legume Lupinus albus var. multolupa, raw or after α-galactoside extraction, and its effect on the bioavailability of protein, P, Ca, and Mg by growing rats wasevaluated using a balance technique. The protein and dietary fibre content of the lupin floursstudied was high, and 89–94% of the dietary fibre was present as insoluble dietary fibre. The α-galactoside extraction process did not disrupt the nutritional quality of protein, and the digestive and metabolic utilisation of this nutrient was high and comparable with that obtained from a casein–cystine control diet (pair-fed to the average daily food intake of the experimental groups fed the different lupin diets). Bioavailability of P, Ca, and Mg from the lupin diets tested was high, and supplementation of an exogenous microbial phytase (750 phytase units/kg) did not cause any further improvement. Mineral content in the bone tissue (femur and sternum) did not correlate to mineral balance, which, on the other hand, was related to the mineral content of other tissues such as blood, plasma, liver and kidney. Due to its ability to grow under adverse edaphic and climatic conditions and to its good nutritional quality, α-galactoside-free lupin flour supplemented with the required amounts of minerals and vitamins to meet nutrient requirements can be used as an excellent dietary source for the preparation of dietetic products.

The aim of the present work was to investigate the effects of trans-10,cis-12conjugated linoleic acid (CLA) on the activity and expression of lipogenic enzymes and lipoprotein lipase (LPL), as well as on the expression of transcriptional factors controlling these enzymes, in adipose tissue from hamsters, and to evaluate the involvement of these changes in the body fat-reducing effect of this CLA isomer. Thirty male hamsters were divided into three groups and fed atherogenic diets supplemented with 0 (linoleic group), 5 or 10g trans-10,cis-12CLA/kg diet, for 6 weeks. Body and adipose tissue weights, food intake and serum insulin were measured. Total and heparin-releasable LPL and lipogenic enzyme activities (acetyl-CoA carboxylase (ACC); fatty acid synthase (FAS); glucose-6-phosphate dehydrogenase (G6PDH); and malic enzyme (ME)) were assessed. ACC, FAS, LPL, sterolregulatory element-binding proteins (SREBP-1a), SREBP-1c and PPARγ mRNA levels were alsodetermined by real-time PCR. CLA did not modify food intake, body weight and serum insulin level. CLA feeding reduced adipose tissue weight, LPL activity and expression, and increased lipogenic enzyme activities, despite a significant reduction in ACC and FAS mRNA levels. The expression of the three transcriptional factors analysed (SREBP-1a, SREBP-1c and PPARγ) wasalso reduced. These results appear to provide a framework for partially understanding the reduction in body fat induced by CLA. Inhibition of LPL activity seems to be an important mechanism underlying body fat reduction in hamsters. Further research is needed to better characterizethe effects of CLA on lipogenesis and the role of these effects in CLA action.

Soya isoflavones are thought to be cardioprotective due to their structural similarity to oestrogen. In order to investigate the effect of soya isoflavones on markers of endothelial function we conducted a randomised, double-blind, placebo-controlled, cross-over study with thirty healthy postmenopausal women. The women consumed cereal bars, with or without soya isoflavones (50mg/d), for 8 weeks, separated by an 8-week washout period. Systemic arterial complince (SAC), isobaric arterial compliance (IAC), flow-mediated endothelium-dependent vasodilation (FMD) and nitroglycerine-mediated endothelium-independent vasodilation (NMD) were measured at the beginning of the study and after each intervention period. Blood pressure (BP) and plasma concentrations of nitrite and nitrate (NOx) and endothelin-1 (ET-1) were measured at the beginning and end of each intervention period.NMD was 13·4 (sem 2·0) % at baseline and 15·5 (sem 1·1) % after isoflavone treatment compared with 12·4 (sem 1·0) % after placebo treatment (P=0·03). NOx increased from 27·7 (sem 2·7) to 31·1 (sem 3·2) μm after isoflavones treatment compared with 25·4 (sem 1·5) to 20·4 (sem 1·1) μm after placebo treatment (P=0·003) and a significant increase in the NOx:ET-1 ratio (P=0·005) was observed after the isoflavone treatment compared with placebo. A significant difference in SAC after the isoflavone and placebo treatment was observed (P=0·04). No significant difference was found in FMD, IAC, BP and ET-1. In conclusion, 8 weeks' consumption of cereals bars enriched with 50mg soya isoflavones/d increased plasma NOx concentrations and improved endothelium-independent vasodilation in healthy postmenopausal women.

Reactive hypoglycaemia (RH) is a condition that has been popularised in the media and lay literature, particularly that targeting women, over the past 30 years. The objective of the present study was to investigate whether a non-patient group reporting symptoms that they attributed to a low blood glucose level would demonstrate biochemical hypoglycaemia when symptomatic and whether their habitual diet and activity level differed from those of controls. Thirty non-obese, healthy women (aged 19–45 years) reporting symptoms more than once a week that they attributed to hypoglycaemia (RH group), and eighteen controls, measured their finger-prick blood glucose level 3h after breakfast and lunch, and recorded their diet and activity daily for 7d. The RH group also measured their blood glucose when symptoms were being experienced. Symptoms less than 4h after eating were classed as postprandial. The mean postprandial blood glucose level in the RH group when asymptomatic (4·66 (sem 0·08) mmol/l) was significantly lower than that of controls (5·05 (sem 0·11) mmol/l; P<0·01). Symptoms occurred 2·6 (sem 0·13) h after eating, at a lower blood glucose level (4·18 (sem 0·10) mmol/l; P<0·001) than when the women were asymptomatic. On symptomatic days, the RH group were more physically active than the controls (1·64 (sem 0·04) v. 1·50 (sem 0·03) multiples of resting energyexpenditure; P<0·05), with a lower energy intake (7901 (sem 311) et al.. 9332 (sem 227) kJ; P<0·001). In conclusion, subjects reporting symptoms they associated with hypoglycaemia generally did not demonstrate biochemical hypoglycaemia but did have significantly lower blood glucose levels than controls. Higher physical activity and a failure to match energy intake to estimated energy requirement may be important in the aetiology of symptoms.

Zn has been shown to possess antioxidant properties in vitro and in vitro. As inadequate dietary Zn intake has been reported in these populations, Zn supplementation may protect against oxidative stress and thereby limit the progression of degenerative diseases in such populations. We conducted the present study to evaluate the long-term supplementation effects of two moderate doses of Zn on in vitro Cu-induced LDL oxidation in French men and women.Three groups of sixteen healthy subjects aged 55–70 years from each sex participated in this randomized double-blind, placebo-controlled study. Each group received for six months either 0, 15 or 30mg supplemental Zn per d. At the beginning and at the end of the supplementation periods, dietary intakes of Zn, Cu, Fe and vitamin E were estimated using 4d food-intake records (including the weekend) and the GENI program. Zn, Cu, Fe and vitamin E statuswere also determined. In vitro LDL oxidizability (basal conjugated diene level, maximal conjugated diene formation and lag time) and lipid parameters were also determined. Dietary intakes of Zn, Cu, Fe and vitamin E were adequate in this population. Zn supplementation significantly increased serum Zn levels but did not significantly modify Cu, Fe or vitamin E status. However, Zn supplementation had no effect on in vitro LDL oxidation parameters, nor were there any sex-related differences in in vitro LDL oxidizability. The present study showed that long-term Zn supplementation of healthy subjects aged 55–70 years had no effect on in vitro Cu-induced LDL oxidation under the study conditions.

Fructo-oligosaccharides (FOS) are widely used in commercial food products. Most studies on FOS concern the health benefits, but some negative effects were recently reported concerning thefaecal cytotoxicity and excretion of mucin-type oligosaccharides in combination with a Ca-restricted diet. The present study was performed to investigate whether these effects of FOS are observed in adults consuming a regular diet unrestricted in Ca. The study was a randomised, double-blind, placebo-controlled crossover trial, involving eleven healthy adults, who consumed 25–30g FOS or maltodextrin (control) in a random order for 2 weeks in addition to their regular diet. Stools were collected for analysis of pH and SCFA (as markers of fermentation), for the assessment of faecal water cytotoxicity, and for the analysis of alkaline phosphataseactivity (as a marker of epithelial cell turnover) andO-linked oligosaccharides (to estimate the excretion of mucin-type oligosaccharides). FOS consumption significantly altered bacterial fermentation (increased percentage of acetate, decreased percentage of butyrate) and tended to decrease stool pH. Furthermore, FOS consumption resulted in a significantly higher stool frequency and in significantly more complaints of flatulence. No significant differences between the control and FOS period were observed in the mean cytotoxicity of faecal water (37·5 (sem 6·9) % v. 18·5 (sem 6·9) % P=0·084), in mean alkaline phosphatase activity (27·7 (sem 2·9) v. 24·6 (sem 3·2) U/g dry faeces; P=0·496) or in the mean excretion of mucin-type oligosaccharides (49·9 (sem 4·0)v. 53·5 (sem 4·3) mg/g dry faeces; P=0·553). We conclude that dietary FOS in a dose up to 25–30g/d altered the bacterial fermentation pattern but did not affect faecal cytotoxicity or the faecal concentration of mucin-type oligosaccharides in human adults consuming a regular diet.

Phyto-oestrogens, naturally occurring hormone-like chemicals in plant food, may play a protective role against hormone-related chronic diseases. South Asian migrants in the UK have a lower incidence of hormone-related cancer than their hosts but the extent to which this difference may be due to phytoestrogen intake is not known. The aim was to compare habitual phytoestrogen intake in first-generation South Asian migrant women and native British women. South Asian (n 221) and native British women (n 50) were recruited from general practitioner lists and were asked to provide monthly 24h recalls for a period of 1 year. An enhanced phytoestrogen database was compiled using data from a literature search and unpublished data. A sub-sample of South Asian women (n 100) and the native British women (n 40) also provided blood samples every 3 months during the 1-year period. The median daily intakes (μg/d) of isoflavones (184·2 v. 333·9) and lignans (110·8 v. 148·8) were significantly lower in South Asians than in the native British (P<0·001, P=0·04 respectively).There were no significant differences in mean plasma isoflavone levels (nmol/l) but plasma enterolactone was significantly lower in the South Asians (13·9 (sd17·5) v. 28·5 (SD23·3),P<0·001). The main sources of phytoestrogens were bread and vegetables in both ethnic groups. Habitual phytoestrogen intake in South Asian and native British women was below 1mg/d and was higher in the native British diet. The present study does not support the hypothesis that differences in phytoestrogen intake, or in circulating levels, could explain differences in hormone-related cancer risks between these two populations.

Essential fatty acids (EFA) have proved to be important for normal bone mineral density (BMD) and bone growth in animal studies. Patients with cystic fibrosis often have low serum EFA levels, and low BMD has also been reported in patients with normal anthropometry. The aim of the present study was to analyse if BMD during a 2-year period was related to fatty acid status in patients with cystic fibrosis. Fifty-four patients, aged 6–33 years, were studied prospectively. BMD was measured with dual X-ray absorptiometry, and fatty acid concentrations in serum phospholipids were determined with capillary GLC. The cystic fibrosis patients showed low linoleic acid concentration and a high arachidonic acid (AA):DHA ratio in serum. The high eicosatrienoic acid:AA ratio, an indicator of EFA deficiency, increased further over 2 years, as did the total concentration of saturated fatty acids. In the adults there were no significant changes in fatty acids during the study. In the children, positive correlations were found between palmitic acid and bone mineral content in the lumbar spine and femoral neck. The lumbar spine BMD Z score correlated negatively with the AA:DHA ratio. No correlation was seen in adults except for a positive correlation between EFA deficiency index and the areas of lumbar spine and femoral neck. The present results imply that fatty-acid status influenced BMD in cystic fibrosis children, but not in adults, indicating that fatty-acid status wouldbe important for bone growth.

Recommendations for daily energy requirements use factorial calculations based on BMR. Expressing total energy requirements as a multiple of BMR is based on the assumption that BMR is equal to overnight metabolic rate (OMR). The objective of the present study was to determine if BMR is an appropriate proxy for OMR in children, young adults and elderly. Data are presented of thirty children (11 (sd 2) years), thirty young adults (25 (sd 5) years) and fifty-nine elderly(61 (sd 5) years). OMR was measured in a respiration chamber while sleep was not hindered and BMR was measured directly afterwards with a ventilated hood system under strictly controlled conditions. The mean ratio of OMR:BMR was 0·92 (sd 0·09) for children, which was significantly different from 1·00 (P<0·001), 1·00 (sd 0·07) for young adults and 1·06 (sd 0·09) for elderly whichwas also different from 1·00 (P<0·001). Foradults, BMR is an appropriate measure of OMR. In children, the use of BMR to estimate OMR wouldintroduce an overestimate and for elderly an underestimate.

Consumption of soya-based nutrients is increasing in modern society because of their potentially protective effects against chronic diseases. Soya products are also heavily advertised as alternative drugs for relief from symptoms of the menopause and for hormone replacement therapy.However, because of their oestrogenic activity, negative effects of isoflavones have been postulated. Therefore, we analysed influences of soya isoflavones, major soya constituents with endocrine activity, on thyroxine (T4) binding to its distribution proteins. Serum binding of 125I-labelled l-T4 was analysed in the absence or presence of increasing concentrations of soya isoflavones using non-denaturing PAGE for analysis. Complete displacement of [125I]T4 binding to transthyretin (TTR) was observed in human serum incubated with genistein at concentrations >10μm; interference started at >0·1μm. Glycitein showed decreased and daidzein the lowest displacement potency. [125I]T4 was displaced to albumin in rat and to T4-binding globulin in human serum. Soya isoflavones also obstruct [125I]T4 binding to TTR in human cerebrospinal fluid (CSF). The inhibitory effect was confirmed in direct binding assays using purified TTR with 50% inhibitory concentration values of 0·07μm for genistein, 0·2μm for glycitein and 1·8μm for daidzein. The present study underlined a potent competition of soya isoflavones for T4 binding to TTR in serum and CSF. Isoflavones might alter free thyroid hormone concentrations resulting in altered tissue availability and metabolism. As a consequence of this interference, one could expect a disturbance in the feedback regulation of hormonal networks, including the pituitary–thyroid–periphery axis during development and in adult organisms.